The Age of Hyper Acceleration: AI, AGI & Beyond! | Josh Kale
01:04:43 Share

The first person to break down and like reverse engineer the first protein, it took him 12 years to do. And he took this protein and he crystallized and he shot it with x-rays and then he actually used like a ruler and pencil to like connect them together and kind of reverse engineer this. And then over the course of the next 60 or so years, we were able to finally discover 150,000. And now because of AI, we've just discovered 250 million. 150,000 to 250 million, quite a different number. And that 250 million...

was only over the course of like the last two years and change. It was very quick. Okay, so David, there's this quote by Steve Jobs that I love that says, "Everything around us that we call life is made up by people no smarter than we are." And if you think about it, it's true. Like the clothes on our back, everything around us

No one smarter than us has ever made anything that we have here. And the same is true if we go back like 40,000 years to cavemen. We're not really even much smarter than they were. We have the same brain size, roughly same cognitive ability. We benefit though from the collective accumulation of knowledge over time. So it's like,

this snowball where theirs was very small, but over time it's kind of rolled down this hill faster and faster and it's grown. And now we benefit from this collective knowledge, but we're not actually smarter. So the question I think we want to answer in this podcast and the really interesting question is like, what happens to the world around us when people are actually smarter than we are? Like what happens when they're 10% smarter, 20%, 100, 1000% smarter? And we could ask that question because this new thing called AGI, which is

Artificial general intelligence, it kind of has this fuzzy explanation, this fuzzy definition. My definition of it is like a form of intelligence that is smarter than a human across pretty much every medium.

And the initial expectations from experts were the 2040s, 2050s, 2060s. But the reality now is that we are months away from this intelligence instead of decades. So it's like, how crazy is this world about to be when everything around us is actually made up of people that are much, much smarter than we are? Hang on, let's back up with this. So...

I think you're setting this foundation of humans have always been on this exponential growth curve because we have capable brains. That's why we're humans. And we have been like building technologies, building tools. We created the wheel and then everyone had the wheel and then we could create like the next technology after that on top of the foundation of having the wheel. And then we created, you know,

plumbing, agriculture, all these technologies got layered onto each other. And that's where we have this notion of like 2% growth year over year. And 2% growth does compound. Like that is an exponential growth curve. If you're growing 2% every single year, you are growing up an exponential. But I think what you're doing, what you're setting us up for this episode, Josh, is you're

calling for an inflection point, a change in that growth curve because something different is happening. Where previously we have just been layering technologies on each other and humanity has been accelerating. Once we got the internet, we're going a lot faster because we somehow filled in all the gaps between building the wheel and having the internet and we are layering all of our human advances on top of those previous technologies. But you're saying that this is different, right?

In addition to all of that compounding growth curve, we are getting this new technology that is materially changing the rate of growth because where previously all of the intelligence of humans who created the wheel to create the internet was about the same. We were all running on the same wetware. Our brains were about the same level of computational capacity. And everyone has been like...

able to become about the same level of smartness about each other. That's what you're saying is like the history. That's history up to today. And you're saying today is different. Now it's different. That's what you're saying? This time it's different. And it actually is different. And I have this chart that I pulled up on the screen that I'll try to describe to people, which is the chart of human progress in relation to time over a long period of time. And it mostly looks like a fairly flat chart, except there is this like slight incline that starts to grow slightly exponentially about

60 years ago. And that's when the invention of Moore's law happened. So this was a big moment for us where we had transistors and transistors led to computers and computers led to a whole bunch of technology that accelerated things pretty quick. So we have now this steeper ramp up. And then today things change. And the reason I say this time it's different is because we have this new law called Huang's law. And it's loosely debated on how accurate it is, but it is basically the speed of training a cluster

It improves 25-fold over every five years. And Moore's law was that the amount of transistors on a chip doubles every 24 months. So this is a 5x increase in Moore's law that initially started this growth trajectory. And that's why I think the chart will start to look like this.

And it's really hard for the people that are listening. It's basically vertical. The line goes straight up. It's an exponential curve with a pretty strong kink in it. It's still curving upwards, but it goes up much faster, much sooner. Yes. And the idea is that like this is going to get pretty fast because we're entering the age of hyper acceleration. And the reason for that is because

Like we said, everything has been built on this wetware that is like pretty dumb. Like we're smart, but we are only so smart relative to what's coming. Now, since we're able to accelerate five times faster than we were with Moore's law in terms of training this new form of intelligence, there's going to be a lot of weird and wacky things that start to happen really, really quickly. And we're kind of seeing this. It's similar to COVID where like humans are very durable and very malleable and we're good at getting over trauma. So you can kind of like,

mute out your emotional bands to deal with this stuff. But if you just look at what happened in the last week, we were able to give gene therapy to a blind Irishman. Gene therapy to a blind Irishman. What does that mean? What happened? So we can inject these new forms of genes into a person who's blind and give them eyesight back. That

That's pretty cool. We made a blind person see again last week. Yes. And that's just through gene therapy. That's not through neural interfaces. We fixed blindness in one individual. In one individual, just like last week. And then there was another breakthrough with ARC EVO 2.0.

DNA sequencing model. So there's this really interesting thing that's happened with DNA and we won't go too deep, but basically if you can reverse engineer a protein and rebuild it, you can create all these new forms of technology. You can solve forms of cancer, you can cure forms of Alzheimer's. We discovered 250 million of those when over the past 60 years, we've only discovered 150,000 and the first person, it took him 12 years just to discover one. So we now have this like crazy slew of technology that was just released last week.

Google AI has a co-scientist that accelerates the development of science. You can query 100 PhD students' worth of compute data

for about like 10 cents in 10 minutes. Microsoft has a topological quantum chip. We have all these new humanoid figures. Croc 3 has a new AI model. There's all these new machine learning models that are frontier models. It's like accelerating very, very quickly. And every week there's something new. Just yesterday, for the people that remember the DeepSeek R1 model that broke Wall Street, that came out only two months ago. Yesterday, Alibaba released QWQ, which is 20 times more efficient than DeepSeek R1. And

slightly better. So already the thing that broke Wall Street has been broken 20 fold in a matter of two months. So things are going super fast. We're developing everything super fast and that's because it's all kind of downstream of intelligence is the smarter we are as people, the more complicated problems we can solve

And the faster we can solve them and the faster we can accelerate these new problems worth solving. Let's back up and define Moore's Law and extrapolate that into Huang's Law once again. Moore's Law, my understanding of Moore's Law was it was just really an extrapolation on the computational power of a CPU, of a computer chip. And the idea here is that there's this trend line

around how fast a chip can be. And I think you defined it as two Xs every two years. And Moore's Law has been in effect for decades, many, many decades. My understanding of it is like it runs up against a wall based off of like the thickness of silicon down to the nanometer level. And so we eventually run out of slack to grow in Moore's Law. Like my last understanding of Moore's Law is we are approaching that limit. Maybe you can update me on that. But then now we, in addition to that, we also have Huang's Law, which is,

It's not apples to apples. It's more apples to oranges, but it's of the same spirit where it is talking about the computational power as it relates to, I mean, I'm guessing Huang is in Jensen Huang from NVIDIA. Jensen Huang, yes. Yeah, so it's the computational power of clusters that is largely computing AI stuff.

Maybe you can fill in the gap of all the gaps I just left. Yeah. So Moore's law is the amount of transistors on a chip doubles every two years. That is roughly like you could think of it as the computer processor, the CPU, the brain of a computer doubles in speed every two years. Huang's law is slightly different. Apples to oranges, largely debated. But Huang's law is every five years, we get a 25x improvement in training capability for GPUs, which means we can train these LLMs, these AI models,

25 times faster every five years. And Moore's law is running into this interesting law where it's actually constrained by physics now, where the particles have gotten so small that we're not quite able to fit many on there without these new innovations. And I think that's kind of why you're seeing things start to stagnate slightly, but equally an opposite takeoff in the world of AI. And now it's

It's not certain that we won't hit a wall very quickly with Huang's law as well with GPU training clusters, but there's this other side of it with software and the software can actually compensate for the lack of hardware in the case that this ever does degrade. What we just spoke about recently with deep seek and QWQ is they were able to take these giant foundational models that took

hundreds of thousands of GPU clustered together, tons of energy, tons of compute, and they were able to distill them in a model that can actually run on a desktop computer. And that improvement is like a thousandfold improvement. And that happens on the software side of things. So there are these two pillars that are advancing really quickly. There's the software stack and there's the hardware stack, and both of them are going exponential. So even if Wang's law is half of what he says it is, or half of what is expected, the

opposite forces that are happening in the software field will continue to accelerate that even faster. So we have these two tailwinds both going super quick. Right, right. Okay, so Hoang's Law is a 25x of computational efficiency every five years, which is pretty damn steep, much deeper than Moore's Law. But then you also gave us the 20x performance or efficiency increase between the Alibaba model that got released earlier this week versus the DeepSeq R1 model that got released two

two months ago. So we're layering a 25x every five years of computational power. You're multiplying that with what we are seeing in the efficiency gains of the highly competitive AI lab model releases. And that's software, right? And I love this metaphor. I use it many, many times on Bankless. There's two ways to scale. You can just

have better hardware. Your Xbox 360 can upgrade to an Xbox One. You have a better chip in there. The hardware is better. But then also, the software devs, the devs writing the games, can write more efficient code. It can use the same hardware more efficiently, create more beautiful games. There's two ways to do this. And we're seeing both happen

At these crazy scales, a 20x increase in efficiency for a model in two months is insane. That growth curve can't be sustainable. I suspect it tapers out by the end of this year. And I think it just illustrates that there's a lot of slack in the AI model development race. There's a lot of optimizations to be had. But the fact that we are 20x-ing in efficiency in two months in terms of one model release to the next just tells me that there's so much slack.

optimizations to be done here. And so we actually don't really know where the efficiency gain happens on the software side of things. And also a part of this to take into account is these efficiency gains are happening and they're compounding because of the output of the efficiency gains. So now because we have smarter models to help us, we are able to build even more efficiencies, even more improvements,

And the hope is that it will increase systematically, where as we build these smarter models, they can help us create more efficiency, help us design more efficient chips, and it will be the self-fulfilling acceleration curve. Oh, so the models help us design better hardware, the hardware gets better, and we can run more powerful models, and then the models can help us design more hardware, and it kind of just

naturally converges the optimization point extremely fast. Exactly. Yes. And as we accelerate faster, there will be even more improvements to the world of atoms instead of just bits. So now it can teach us how to build these machines to build this new computer to build this new chip that will make things even faster. So it's kind of this like self-fulfilling cycle up

to wherever we end up going. Okay, and so that's why you're illustrating this kink here, again, on the screen that we're showing, where there's been this compounding growth of humans and the technology that we've had ever since we invented the concept of technology and innovation and science. And there's been just, you know, call it 2% growth year over year over year. We've been on an exponential growth curve. It's been a modest 2%.

But this is why you're saying that there is a kink here and we are entering the age of hyper-acceleration where, sure, we've had more useful tools and we've invented power tools for building better homes. We've invented computers and that really helped accelerate things. But nowhere have we actually cheapened

the cost of intelligence. And the idea here, I think right now, ChatGPT and OpenAI, they're spending a ton of money on just running inference, running people's queries into ChatGPT, and that costs a bunch of money. But also, the amount of inference that they're doing is also quite high. And I think what you're saying here is,

Basically, we are collapsing the cost of intelligence from being expensive, which is like you need to produce a whole entire human. You need to put them through school. You need to put them through college. You need to put them through a PhD program. That takes a lot of time. That takes a lot of energy. That's very costly.

And that was the previous model. And now all of that intelligence inside of the span of five years, 10 years, it's becoming free. Intelligence, like PhD level intelligence will be free in a few short years. Is that what you're saying? Yeah, totally. And that was kind of the essence of why I created this like little visual is because it was mostly the intention of exploring the downstream effects of intelligence.

intelligence as the price rapidly decreases to zero. Because again, everything around us, it requires intelligence to make. But what happens if that intelligence becomes so cheap that it is readily accessible and 10 times smarter than we are by anyone in the world? Like it makes you think a lot about the productive output that we can unlock. And that was the essence for the very steep curve.

I see that there's two sides of this conversation. There's so much optimizations left to be done on the hardware side. There's so much optimizations, way more optimizations left to be done on the AI model side. These things are going to get so much more efficient. The cost of intelligence is going to drop to zero. There's so much optimizations to be done on just like the mere intelligence side of the equation.

But that's the producing intelligence. And then there's the application of intelligence. And that's where you went through some of this. Let's hear it. You talked about the gene therapy, giving a blind man sight for the first time, or like Google AI co-scientists, literally just accelerating science broadly. And so while we have the AI lab wars happening,

the cost of intelligence down to zero. We are still like in the very early days of actually seeing the application of that intelligence. And I think also something that you're saying is like, well, actually some people are applying this and they're applying it into fields all across science, all across academia, all across knowledge. And it's kind of like slipping under the radar. But if you pay attention, you'll kind of notice that we're doing some pretty crazy things really quickly.

Yes, that's absolutely right. One of the more interesting ones that I do want to highlight here because I think it's so cool is the DNA sequencing and the protein sequencing. That is like a core foundational part of biology. Of life. The first person to break down and reverse engineer the first protein, it took him 12 years to do. And he took this protein and he crystallized and he shot it with x-rays and then he actually used a ruler and pencil to connect them together and kind of reverse engineer this. And then over the course of the next...

60 or so years, we were able to finally discover 150,000. And now, because of AI, we've just discovered 250 million. 150,000 to 250 million. Quite a different number. And that 250 million was only over the course of like the last two years and change. It was very quick. And

And this was also, relatively speaking, two years ago. It was very early in the AI cycle. So by using this AI, we were able to unlock this entire foundational core of biology. And when you could break down proteins to the simplest form, it unlocks a lot of really cool innovation. So like I mentioned earlier, Alzheimer's is a direct result of protein misfolding. Same with a lot of forms of cancer. But there's also these other applications outside of just biology where you can form a protein structure that eats plastic and you could just

send it out into an ocean and design it so that it doesn't harm the ecosystem but it can just eat a lot of plastic and convert it into something useful. Or you could design these new materials that are much more heat resistant like a rubber tire that never popped, that lasts forever. Material science starts to expand and there's this whole world of biology that we've just unlocked that now we can manipulate and mutate for the first time because we understand it. So it's starting to give us these really like foundational bases of knowledge of

of biology that now we can take and we can use and we can apply to the real world. And that unlocks an entirely new industry. I think the subject that you're talking about is called synthetic biology. At least that's how I understand it. And maybe the best way to explain this to listeners is the way a computer works, like going back to a simple Turing machine, is you have this tape.

And you have this tape of like serial numbers. It all goes down to like ones and zeros, but like the way a computer works is they processes the string of digits serially. And then we just make faster computers to do these things faster. And then we also have multiple cores to do them in parallel.

But ultimately, at the end of the day, it boils down to a serial string of characters, a serial string of bits that a computer processes. And DNA is that same kind of structure. It is a serial chain of proteins with more combinations than just ones and zeros. It has the A, T, G, E DNAs. But ultimately, it's just a serial string of information. And it's kind of like the organic code for biology.

And there's this whole universe out there where like, okay, if we figure out how to string the correct order of proteins together, we can make some really cool things. And I think what you're doing here, Josh, is you are combining like intelligence, complete and comprehensive intelligence onto like how we can correctly order biology to create any sort of like organic structure that can do almost anything that we want. We can cure blindness. We can produce this organism that consumes plastic that like helps fight

fight climate change, whatever. And because it's biology, the actual applications are pretty boundless. They can kind of touch anything. And that is just synthetic biology. There's still like all the other subjects that we have to get into, which is like...

rocketry, like life on Mars, anything else, like any other science that any listener wants to imagine in their head. And so this is why you're saying we're entering the age of hyper-acceleration because every single industry is about to hyper-accelerate. Yes. Everything around us that was made by us, if we were smarter, can be improved. So you will see these improvements across the board. Another interesting one about the synthetic biology too is the entire food supply chain and how we make food, how we create food, how food

gets genetically modified over time. All of that changes too. So there's so many of these different areas that stand to change so much once we understand them better and we know how to refine them better to get better outputs. - Okay, so between the eras of

AI labs becoming hyper-efficient, creating massive intelligence, dropping the cost of intelligence to zero. And then before we get to the actual changes in the industries that are actually where the rubber meets the pavement, we're going to have kind of like the investment layer or the economics layer. This is going to change the economics of everything. Investors are going to reallocate capital.

The cost of things are going to change and fluctuate, I think go down. Before we actually talk about, and talking about all the possible ways that this impacts us is literally- Oh, we'd be here forever. We'd be here forever. We'd just be talking about like the future of the universe. So like, let's talk about what's the economic impact of this in the short term. It probably leads to something

some sort of rapid deflation in the type of industries that it affects. So just to define deflation versus inflation, a lot of people know what inflation is. It's just kind of like the price of goods and services slowly increasing over time. They're very familiar. They see that in the grocery store. We see that gas stations, everything you buy has slowly increased over

over time. And the reasoning for that is there's this really great contrarian take by Peter Thiel that I love where he says that we haven't actually accelerated a whole lot outside of the world of technology. Meaning if you took your grandma and you just kind of froze her in time 50 years ago and then dropped her into your living room today,

not much would look very different. If I look out the window, I'm looking at New York City, I'm looking at bridges and buildings, and those have all been there for the last 50 years. The roads, the infrastructure, a lot of it has kind of been the same. - The cars look different, but they're still cars. - They look different, but like they still run on the same fuel. They still have the same materials that they're built with. They just have computers inside and the computers make them a little bit smarter. But outside of the world of computers, we haven't actually accelerated very quickly.

So we haven't had this like gross overproduction surplus of things that we need that can lower the price. So we generally have seen inflation there, but the place where we have seen deflation is in technology. And that TV on the wall will look like magic to her and the phone in your pocket where you can reach anything in the world instantly at your fingertips. That seems like magic. And the cost of that has come so low that now it's accessible to basically anyone.

So that is deflation. That's where we see it in technology. And it's probably why we haven't seen it in many other industries outside of it. But once we're able to apply this new intelligence to these new industries to get that surplus, hopefully the downstream effects of that are more abundance, more surplus, and a lower cost of goods, a higher quality of living. I think we'll start to see that once we start to manufacture more in the world of atoms versus bits, which is digital. And that, Ben Klisnach, is where we are going to take this conversation next.

Welcome to Bankless, where we explore the frontier of internet finance and internet money, and now the future of Adams, too. This is David Hoffman. I'm joined here once again with our very own Josh Kale. Josh, welcome back to Bankless, your very own podcast. Thank you. Oh, it's great to be here again. I'm so grateful that I have someone who is interested in hearing me talk about

this because this is what I think about all the time. So thank you for having me again. It's a pleasure. Yeah, I really enjoyed the last conversation that we did. If this is your first time listening to Josh, we did one other podcast, the one last week, really just kind of doing the landscape of the AI arms race in a very like zoomed out

way where like sure it's fun to talk about like the competitiveness between these different ai labs but i think josh really synthesizes information in this very big way and that's what we want to do here again on this episode so we're going to continue this episode but first we're going to talk to some of these fantastic sponsors that make this show possible

In the wild west of DeFi, stability and innovation are everything, which is why you should check out Frax Finance, the protocol revolutionizing stablecoins, DeFi, and Rolex. The core of Frax Finance is FraxUSD, which is backed by BlackRock's institutional Biddle Fund. Frax designed FraxUSD for besting class yields across DeFi, T-bills, and carry trade returns

all in one. Just head to Frax.com, then stake it to earn some of the best yields in DeFi. Want even more? Bridge your Frax USD over to the Fraxdal Layer 2 for the same yield plus Fraxdal points and explore Fraxdal's diverse Layer 2 ecosystem with protocols like Curve, Convex, and more, all rewarding early adopters. Frax isn't just a protocol, it's a digital nation powered by the FXS token.

governed by its global community. Acquire FXS through Frax.com or your go-to DEX, stake it, and help shape Frax Nation's future. Ready to join the forefront of DeFi? Visit Frax.com now to start earning with FraxUSD and staked FraxUSD. And for Bankless listeners, you can use Frax.com/r/bankless when bridging to Fraxel for exclusive Fraxel perks and boosted rewards. Introducing Unichain. Built for DeFi and powered by Uniswap, Unichain is the fast,

decentralized Layer 2 designed to tackle blockchain speed and cost challenges. With this mainnet now live, you can enjoy transactions at up to 95% cheaper than the ETH Layer 1, all while benefiting from an impressive one second block time that will be getting even faster very soon. Unichain is the first Layer 2 to launch as a stage one roll up

on day one. That means it comes with a fully functional permissionless proof system from the start, increasing transparency and further decentralizing the chain. More than 80 apps are joining the Unichain community, including Coinbase, Circle, Lido, Morpho and Uniswap. You'll be able to bridge, swap, borrow, lend and launch new assets and more from day one. Built by Uniswap Labs, the team behind the protocol that's processed over 2.75 trillion in all time volume with zero hacks, Unichain truly enhances DeFi experiences with faster, cheaper and seamless transactions even across chains.

and soon the Unichain validation network will allow anyone to run a node and earn by securing the network. Visit uniswap.org and swap on Unichain today. The Arbitrum portal is your one-stop hub to entering the Ethereum ecosystem. With over 800 apps, Arbitrum offers something for everyone.

Dive into the epicenter of DeFi, where advanced trading, lending, and staking platforms are redefining how we interact with money. Explore Arbitrum's rapidly growing gaming hub, from immersed role-playing games, fast-paced fantasy MMOs, to casual luck battle mobile games.

Move assets effortlessly between chains and access the ecosystem with ease via Arbitrum's expansive network of bridges and onrifts. Step into Arbitrum's flourishing NFT and creator space where artists, collectors, and social converge and support your favorite streamers all on-chain. Find new and trending apps and learn how to earn rewards across the Arbitrum ecosystem with limited time campaigns from your favorite projects. Empower your future with Arbitrum. Visit portal.arbitrum.io to find out what's next on your Web3 journey.

Okay, Josh, with the arrival of AI, people like to talk about the job market because there's that classic line of AI is going to come take my job.

And then the usual rebuttal to that is, well, AI might take your job, but the more likely thing, it's actually going to be somebody who's using AI better than you is going to take your job, not strictly AI. And then there's also the additional rebuttal to that. Well, it's like, yes, there might be some job destruction, but it's a creative destruction and there's going to be new jobs that are created.

You know, think about when the internet came and we were worried about newspapers just going out of business, which, you know, definitely happened. Although, you know, there are still, you know, the Wall Street Journal, there's still the New York Times. And so there was some like elimination of jobs as a result of the internet. But then like YouTubers and influencers created this huge economy inside of the internet. And so this idea of creative destruction came where, yeah, we're going to lose some jobs, but we're going to create a lot of new opportunities ahead.

there's one sector that I think people are actually going to just materially lose their jobs and they are going to have to figure out something else to do.

because you're talking about deflation. We're bringing the cost down. Uber's biggest cost is probably paying drivers. And we're already seeing Waymo in San Francisco have just driverless cars. And so I think there are material industries there. There's going to be massive job. You don't really lay off your Uber driver, but that's millions of jobs going away. What do you think about this conversation? Yeah, it's a scary thing. And I want to preface it for the people that

are worried about it with just the essence of solving problems, which is what we're doing. We are solving the problem that transportation is very unsafe. It is very costly and it is very low production. Meaning you have to sit there and you have to drive and cars kill the most amount of people every year. And it's just this very dangerous thing. Like you are allowed as a, what, 17 year old to drive a five ton vehicle at a hundred miles an hour anywhere in the world.

That seems a little scary, a little unsafe. So what's happening here is we are solving problems and there will be this dislocation of people who were solving the past problems. But it's important to note that like solving these problems creates a more exciting world.

When you solve a problem, generally it leads to more problems, but these are better problems to solve, which is why the world around us improves. It's not because we reduce the amount of problems, but because we solve the worst ones, multiplying the amount of better problems to solve. So we've kind of seen this happen with the agricultural revolution where there was seed drilling and crop rotation and crossbreeding and all

all of the people who did those jobs originally, they lost their jobs. But the productive output, the downstream effects of it were so much higher. And it happened again in the industrial revolution where we invented like the steam engine and we invented the production line, textile machines. And the people who did that, they lost their jobs. But the steam engine created a train and then the train created railways and then railways created transportation and cross-country distribution of goods and services. So it opened up all these additional industries. And I think that's probably what's gonna happen

Cars is a great example because, again, very unsafe, very expensive. If you can remove the human element from that, suddenly your one-hour car ride to the airport is a fully productive time where you don't have to drive yourself. You can just do stuff on your phone. You can take a nap. The amount of accidents that will happen will be much lower. It'll be much safer. You don't have to worry about it. And the cost, because there is no human element, will drop aggressively. I think

It's a few dollars currently per mile is the way you can gauge transportation in New York City. That should drop down to...

sense once these cars are able to drive themselves. So it's one of those things where, yes, the cab drivers will lose their jobs, but the world will be a better place because of it. And the hope is that the productive unlock from this new technology will enable even more jobs for them to take and more interesting jobs. I think this conversation is quite timely for me personally, because right after this podcast, I go and on Turo, there's this app out there for people who don't know who can just like rent a car. It's like Airbnb for cars.

So there's this, some guy like four blocks away from me has a Tesla that I'm going to rent and I'm going to drive for six hours upstate. And Tesla has auto driving, but I have to keep my hands on the steering wheel because of regulation. And so that is six hours where I am unproductive. Like my most productive form is going to be listening to a podcast and like growing my knowledge. And that's, that's if I have the brain power for that. So otherwise I'm just like stuck in this car for six hours. It's going to cost me like

Because I'm going for the weekend. It's going to cost me like $300 or $400 to rent this car for the weekend. So it's going to cost me a bunch of money. I'm going to be very improductive for six hours there, six hours back. And I just have to get up there. So there's no other way to do this. And so I think what you're saying, what we're extrapolating here is like, okay, so...

In the future, short-term future, maybe regulations aside, in the short-term future, that $400 cost for a Tesla for three days is actually going to drop to, I don't know, $25, $50. Maybe I don't rent a car, but there's just a car that takes me up there. And during that time, I am not with my hands on the steering wheel. I have my laptop open on my lap. Maybe I'm even doing a podcast. But I'm doing something otherwise productive with my time. I think the low productivity angle of Tesla

you know, Uber drivers, cab drivers is actually really illustrative because like other than getting some other human from point A to point B, that's the only production that actually happens as a result of that job. Yes, that's absolutely right. Actually, I would encourage you to try full self-driving on your road trip up.

because it's really good and it signals how close we actually are to solving this problem. Oh, I intend to do it, but my last time I was in a Tesla, I had to like keep my hands on this earring reel. Now it's gotten better. It's just your eyes need to be on the road. So you don't have to touch anything. You just have to look forward. Can I wear sunglasses? You can, but there's actually these small sensors that can penetrate your sunglasses that can see your eyes beneath them. They've got it pretty good. And this is just really a matter of regulation. The technology is mostly there. In fact, Tesla's planning to roll out their cyber cab network fully autonomous in Austin, California.

in the middle of this year. So we are very close to this happening. And I think a lot of people who haven't tried it don't know precisely how close we are. So give it a try. I'm curious to hear your thoughts, but it's very good and it's very exciting because yeah, you could either take a six hour nap or you could do whatever you would want during that time and do so much more safely than if you had a tired driver or if you yourself were tired.

And that's just the idea of transportation. And I don't know if you have any takes on like airline transportation or if that's getting any cheaper. Oh, so many. Oh, you do? Okay. So this was talking about like car automobile transportation is dropping, call it 100X in cost over when this technology gets absorbed by humanity. Aviation transportation has not progressed at all over decades. It's been one of the most frustrating things in my life. Is aviation transportation cost going down too? Yes, you can apply this to everything. We could have the same conversation about

10,000 different industries. And it will be very similar because once we have better materials, once we have better intelligence, once we have more abundant energy, we can create much better products. So one that I'm super excited about is called Boom Aero. They're a supersonic airplane company. And what they've done is they've been able to basically take the plane and turn it into a supersonic plane that is much more efficient, much faster, and does not disturb the ground level of people with these hypersonic booms. So

their new technology is allowing planes to go way faster. I'm not sure the multiple, but significantly faster at significantly higher volume for the hope is a lower cost. So not only do you get a product that's better, you get an experience that's better, but you actually get a lower cost and higher safety because these,

will be built by really impressive engineers that don't have these edge cases that you've been seeing with these airplanes that are crashing and burning. The airline industry has gotten a little scary and it's mostly because it's a duopoly between Boeing and Airbus. Yes, I believe those are the two. And the

They have no incentive to do right by the people because they are the monopoly. And if you can start to introduce these new competitors through this new technology that provide such a better experience at such a better price, then there's no reason why 10 more of those won't come into the market. And gradually over time, we will just remove the duopoly that is these unsafe kind of crappy airline companies and we'll have hyperinflation.

hypersonic travel and we will have hypersonic travel that's like really comfortable and really cheap and affordable and it's mostly a matter of manufacturing materials and regulation those are the three things that are stopping that from happening but it's happening one thing i think we are doing with this conversation maybe there is something to fill in here is like first we're talking about okay so first we grow hardware and then sophistication of our models that's step one step one grow intelligence step two reduce the cost of intelligence

Step three is three question marks. And then step four is profit, where we have like super cheap transportation, super cheap airlines. And all of a sudden we can do supersonic travel for very, very low. How does that first conversation that we had about AI intelligence actually relate to producing like deflation in transportation costs? Or are we just kind of assuming that eventually with the massive reduction in intelligence that we're going to figure out...

at scale supersonic airline travel. Yeah. So I want to amend your hierarchy slightly. You put intelligence at the top. I think there is one thing that is slightly higher than intelligence and that's energy. To power all these intelligent systems, to power us, we need tons of energy and we don't

currently have enough energy. So I think this all kind of starts with the energy layer. And again, it's synchronous with intelligence. There's this one fact that I love. It's that if you go outside into your backyard, if you go to like a park down the street and you pick up a rock, the rock actually has more potential energy inside of it than the equivalent size piece of coal. And

It's interesting because we use coal, we burn coal, but it turns out that organic plant matter that's been stored underground isn't super dense in energy. Whereas this rock that's down the park, it has trace amounts of thorium and uranium, and they're very low amounts. But the only thing stopping us from extracting that versus coal is understanding how to process it. And the resource thing is really interesting because it's like,

But nothing is really a resource until we assign the knowledge to it that it is a resource. Like we didn't get iron out of nowhere. Some guy found a rock and then he figured out how to smelt it and refine it. And we went from iron ore to iron ingots. Now we have skyscrapers. And there's this kind of progression through resources where as we get more intelligent, we are able to access them more abundantly.

And I think this kind of happens synchronously with intelligence, where the more energy we have, the more we can power these GPU clusters, the more we can understand. And then to get to your question, the downstream effects of it, well, it can teach us how to manufacture better factories. So one of the really interesting developments has been humanoid robots recently. And these are robots that have general purpose intelligence. They have hands, they have actuators that kind of function like humans. And they've been starting to go into factories and starting to manufacture things themselves. And-

a lot of the production line has been simulated through these large AI models that can kind of emulate the efficiencies and inefficiencies of a system and then weed them out for you. So generally with like technology applications or products, the first one is kind of crappy. Like if you bought the first iPhone, it like wasn't that great. The second one was like pretty good. The third one was like really good. The fourth one was amazing. And the fourth one is basically the same as the 16th one. They're all the same because we kind of reached this local maximum. But

If a computer can do all those iterations for you because it's much smarter than you, then the first version can actually be the fourth version, that final version, that local maximum of what we're able to do. So I think that probably applies through most places. As we kind of synchronize this manufacturing machine layer with the intelligence layer, it can basically teach us how to make things. And then our job as humans would be to go and create the infrastructure required to make these things that we want.

to make these hypersonic airlines, to make these self-driving cars, it can remove all the inefficiencies and basically give us the answer, give us the blueprint. - So your equation is that energy plus intelligence equals profit, basically. - Equals profit, yes. - Once we have abundance of energy,

We are currently growing an abundance of intelligence and you can combine those two things and then the universe is our oyster. We can literally unlock every single door once we have those two things. Yes, from down to the biology example that we used all the way up to just like building the most large, like Starship, rocket ships to other planets to go mine other resources. And these things are very abundant. China last week,

thorium in those rocks that I mentioned, they found a deposit of thorium that can power the country for 60,000 years. And it's just sitting there. And they don't really know what to do with it because they don't have the intelligence or the manufacturing capabilities. So they're starting to learn how to create these things called saltbed reactors that are

or safe nuclear reactors to extract it and to refine it, but it's not there yet. And if they had this higher form of intelligence that could feed them a blueprint, like, hey, here's exactly what you need to do in the optimal state to extract this energy, they suddenly have power for 60,000 years. And then they can apply all of that energy to whatever problems they want to solve outside of that. And there's this chart that I have. It's basically showing that there are no energy-poor countries in the world. If you don't have energy, you cannot survive.

produce valuable things. And so that higher energy thing, the higher intelligence thing, they're the self-fulfilling loop and they are upstream of basically all of that profit. Mm-hmm. And we've done a couple episodes with guests like this. One of them is Arthur Hayes where he says he denominates his wealth in hydrocarbons, in energy. Mm-hmm.

And I don't know if he actually does that, but the point stands of just like, well, I mean, what is the dollar really? It's actually the petrodollar. And it's actually petro, the oil reserves of the world, that is actually the fundamental denominator of anything. And like, why is it so valuable? So, well, because it produces...

energy. Josh, have you ever read any of David Deutsch's stuff? Oh man, it's funny. I have the beginning of infinity and fabric of reality like right here. Actually, hold on. Yeah. Okay. Okay. So I think you know exactly where I'm going with this. Josh is zooming over to go get his books. Yeah. Have you read it? I've tried. I've tried. David Deutsch is dense. This is a tough book. I don't want to say I've read it because I have not, but I have tried and I've actually consulted YouTubers on how to go about approaching this book because it is so dense. But

It's very high signal. And I think a lot of people who I trust and respect really love and keep coming back to this book. So that is why I continue to chug away. But yes, I have attempted to read it. You're aware of David George. Okay, so he's got this idea in these books, the two books that you named. He argues that the only fundamental limitation on what humans can achieve is...

is constrained by knowledge. Like all constraints are knowledge constraints at the end of the day. And specifically, it's our ability to like discover the right explanations and create the necessary technology to do things is just a knowledge constraint. And so his idea is that

if something doesn't break the law of physics, then the only thing preventing us from achieving that is the lack of knowledge to get there, right? So space travel, there's nothing in physics that forbids humans from colonizing other planets or traveling between galaxies. The reason why we haven't done it yet is we don't have the knowledge, right? Aging, if aging is just a biological process governed by physical laws, then in principle, we can learn how to reverse it. The only obstacle is the knowledge how.

And so I think this is what you're alluding to is like, okay, we need the energy to power the intelligence. We actually have the intelligence, or at least we're getting it very, very soon. I consider us having the intelligence, but I'm sure we're going to have even more intelligence at the rate of these AI labs competitions. And so once we get the energy, we have the knowledge, and then the world is our oyster. And so like, yes, we didn't really have a clear answer as to how AI actually allows for supersonic travel at

you know, very, very low cost. But it's making this big assumption that, you know, you smash energy and knowledge together and there's literally no problem that we can't access. And that is why you're calling this the age of hyper acceleration and why there's a kink in the compounding growth curve because everything becomes accessible to us by like the end of this decade.

Yeah, really, really soon. And there's this example I mentioned last time, but I really love so much. And it's the mouse in the prime number maze, where you have this mouse that's not super smart. It's in the middle of a maze and it will run forever and never figure out how to get out of it. But if the mouse knows that it's a prime number maze and it understands what prime numbers are, then it can very easily get its way out of this maze. And there's so many questions to be explored.

to unlock that one key bit of knowledge that unlocks this entire world for us. So part of the higher level of intelligence is asking just better questions or even knowing what questions to ask that are worth solving and then pointing it at those questions. But I very much agree with David in the sense that all of this is just an intelligence problem and an energy problem. And if we have unlimited of both,

Or if you have enough resources to harness a seemingly unlimited amount of both, there is no problem that we can't solve. Everything. Like it kind of breaks your mind. In the same way that LLMs kind of broke my mind when I first started using them, the only constraint is your own creativity or your own questions to ask it.

Like, I feel like I'm still not getting the most out of these large language models because I just don't know how to use them quite well. And that's kind of the case with these LLMs as they get better is the hardest thing will be understanding what to ask, what questions are worth learning. Wow, that's deep.

Once we have infinite intelligence at our fingertips, the constraint becomes, what do we ask it? Yeah, within the realms of physics, or perhaps not. Like, perhaps we understand more quantum physics that break all the rules that we have. Like, we can kind of play God. You can create these, like, new forms of babies that are genetically perfect, and they never have any mutations, and they age at the exact rate that you set. Or you could have...

This food that you create that looks nothing like what we've had, but it's the exact nutrient macro complex for your specific body. And it just shows up every day and it tastes delicious. And it's built just for you because we can manufacture it for every person on earth. And this kind of goes across the board for anything, anything and everything that you can imagine. Again, once it's made up by people smarter than we are, it changes everything for everyone.

So it's a weird thing and we're getting there super quick. Can you pull up the graphic about how there are no energy poor rich countries? There are no energy poor rich countries. And so the claim here is that we can look at all of the countries that exist in the world and all of the rich ones, all the ones that are wealthy, have an abundance of energy. Trace over this idea and why this is so important again. Okay.

So what you're looking at is this relationship between electricity consumption and GDP per capita of countries along the world. And you'll notice in the bottom left, there is a lot of countries like Bangladesh, Pakistan, Sudan, Nigeria. These are all energy poor countries in the sense that they haven't quite figured out how to unlock large amounts of energy. And then you reach this threshold, which is kind of set by India and Indonesia, where they're just kind of there. They've consumed, what is that? A thousand kilowatt hours of energy per capita.

And everything above that is the wealthy nations. That's where you'll see in the top right, Norway is actually very wealthy and has a lot of energy. But there you see the United States. A little bit lower down, you see China, Japan. You don't see any of the large dominant countries in the world underneath this threshold that

Because energy is so important, because energy powers our transportation, our food, everything you do on a day-to-day basis requires that, particularly for manufacturing. We have managed to unlock through burning fossil fuels, alternative energy sources like solar, enough energy to power ourselves and to become wealthy.

But we are pretty quickly outpacing our ability to create this energy. And we're starting to see it with data centers where companies like Project Stargate, the $500 billion plan by Masayoshi-san and OpenAI, they need to create

large amounts of power just to power these energies because the grid can't support it. So we do have enough energy. It's very important for energy because everything is downstream of that energy. But again, we are quickly coming up against that threshold of how much energy we actually have versus how much we're going to need.

Looking at this chart, I'm just very much reminded of Y equals MX plus B. Just to be clear, on the vertical axis is energy consumed per capita. And so if you're a country that has a higher energy consumption per individual, you are higher on the vertical axis. And on the horizontal axis is just wealth of the nation.

And so if you're more wealthy nation, you are further on the horizontal axis. And this is just a perfectly linear chart. As in, like, I've never seen something, like, more correlated to, like, wealth than in this chart. It's like, if you have more energy, your country is more wealthy. Like, this has definitely always been a topic of, like, policy debate. And I think this is something that, like, the Trump administration is actually, like, leaning into. It's just, like, more energy, more energy. And that's...

Like the Democrats, the liberals are very hesitant with his type of energy because it's coal. He's into all forms of energy, including coal. The Democrats are like, well, what about global warming? But then the conservative right is like more energy, like accelerate, like innovate, dominate, like let's grow the energy conversation. And politics aside, in terms of strictly like domestic energy,

innovation. What's the energy conversation inside the United States look like right now? Yeah, it's such a shame that technology has become politicized because it is so foundational to the well-being of everybody. So it sucks that there are these split views on how we should go about it. I think currently the ideology is to use as much energy as we can to accelerate quicker, which is something that I'm really excited about. Again, going to the problem solving thing is like,

Solving one bad problem never leads to like a non-problematic state, but it leads to much better problems to solve. So if we do need to burn a lot more fossil fuels, a lot more coal to power these new plants, that will then give us the information we need to unlock nuclear technology, say. That is a big win. And maybe we suffer on a short-term basis where we do produce a lot more pollution, but the second order effects of that vastly outweigh the slowing down and diminishing

diminishing of this accelerative force that will get us out of the problem. So I think currently the United States is roughly aligned with the

energy needs of what we're going to eventually have. There is a move over to nuclear energy, which I think is super exciting because that feels like the end game. That feels like the final form. And it's a shame that nuclear technology has gotten such a bad rap. Again, technology has become this politicized thing. Very few people have actually died from it. It is very safe. The old power plants that did melt down are nothing like these new power plants. And there's this new company actually just last week

I think it's Valor Electronics. I could be butchering that, but they created this modular nuclear reactor that is the size of...

A large apartment, maybe. And the idea is that you can take a stack of these modular reactors and you could place them around a data center like the OpenAI Project Stargate one that's going to require a lot of power. And it can be fully self-sufficient and fully stable with nuclear energy. And I think that is like a really fun direction that we're headed to where we are now able to produce these things. We don't really have the regulatory constraints.

restrictions and safeguards preventing these people from innovating and accelerating on it. And I think that feels like a very healthy direction that I'm super excited about going is now we do have permission to build these new nuclear systems. We do have permission to explore these new forms of energy that are necessary to power this world in which there will be a lot more robotics, a lot more manufacturing than we currently do today.

Imagine a world where your day-to-day banking runs on a blockchain. That's exactly what Mantle is building. Powered by a $4 billion treasury and poised to become the largest sustainable on-chain financial hub. As part of their 2025 expansion, Mantle is introducing three new core innovation pillars that bridge traditional finance with decentralized technology.

First is their enhanced index fund, aiming for $1 billion in AUM by Q1. It provides optimized exposure to Bitcoin, ETH, Solana, and USC, complete with built-in yield opportunities. Next, Mantle Banking promises to revolutionize global value transfer through seamless blockchain-powered banking services, bridging crypto into your daily life. Finally, Mantle X blends AI with DeFi to deliver an intelligent, user-friendly experience for everyone.

And the best part is that this is all in addition to their already launched products like Mantle Network, METH and FBTC. Ready to step into the future of finance? Follow Mantle on X @mantle_official and join the on-chain revolution today. Celo is transitioning from a mobile first EVM compatible layer one blockchain to a high performance Ethereum layer two built on OP stack

with EigenDA and OneBlock finality, all happening soon with a hard fork. With over 600 million total transactions, 12 million weekly transactions, and 750,000 daily active users, Celo's meteoric rise would place it among one of the top layer twos, built for the real world and optimized for fast, low-cost global payments. As the home of the stablecoins, Celo hosts 13 native stablecoins across seven different currencies.

including native USDT on Opera MiniPay and with over 4 million users in Africa alone. In November, stablecoin volumes hit $6.8 billion, made for seamless on-chain FX trading. Plus, users can pay gas with ERC-20 tokens like USDT and USDC and send crypto to phone numbers in seconds. But why should you care about Zello's transition to a Layer 2? Layer 2's Unify Ethereum, L1's

fragmented. By becoming a layer two, Celo leads the way for other EVM compatible layer ones to follow. Follow Celo on X and witness the great Celo happening where Celo cuts its inflation in half as it enters its layer two era and continuing its environmental leadership.

I'm not terribly informed about the nuclear subject, but my loose understanding is that it was one of these industries that, A, got a bad rap downstream of the Cold War. We had the Three Mile Island incident inside the United States. There was a Chernobyl incident in Eastern Europe. And I think that's kind of like marked the tone of nuclear, but it's also just been highly regulated. Like it's been so incredibly regulated that I think that's actually the main culprit as to why we don't see too much

nuclear energy production because we regulated it heavily because of, well, nuclear bombs. And so maybe you can kind of understand that, but it is also, like you say, a shame that we don't have like this robust nuclear industry because again, to my knowledge, just like you said, it's super safe, it's super clean, it's super powerful.

and super abundant. And again, going with this relationship of like more energy equals more innovation, maybe we should like reset our priors on understanding nuclear as an industry inside of the United States. Yeah, the chart was so clear, like the correlation between energy and dominance. It's such a clear cut representation that it's really hard to debate that. It's a shame. And I wish I had these specific examples because I know the meltdowns were not as bad as most people think. But it's a shame that that very early form of technology

impacts so greatly the technology that we could produce today. We have accelerated so much in our understanding of nuclear engineering, but also manufacturing technology, material science, where now we do have these, there are these things called pebble bed reactors and these new gen four reactors that melt down proof and they have a much smaller footprint and they don't lead to any extra pollution and they're self-sufficient and they don't require this grid that's kind of broken down to distribute energy. It could do so locally.

So it largely has been a matter of this bad reputation that it's gotten, but also the people preventing it because they do feel like it's afraid. And we've had in the past this very afraid mindset where we don't want to hurt anybody. We are far too empathetic to harm anybody. But in not hurting anybody, we've harmed everybody because now we have this energy constraint and we don't have the ability to accelerate our way out of it because we have not been allowed to.

So it's very clear like we need energy and we would love more intelligence. These things are amazing and we have the technology to do it without hurting people. We just need to get out of the way and let the people who are ambitious enough to try give it their best shot. Josh, I'm getting a very clear sense that you are a techno optimist. Very much so. A very strong bent towards acceleration. Like you are acceleration ride or die. I don't think you're

not being pragmatic about it. I think you do understand that there are costs, but I think you are just heavily biased towards like, yeah, there's costs and there are solutions to those costs too, which we will also discover if and only if we accelerate. Yeah, it's kind of like religion where I chose this because this one makes me the most excited about waking up in the morning and it feels really good to me. And I'm very familiar with the downsides

and actually less so because I choose to be blindly optimistic. That's not helpful. But I am familiar with the downsides of a lot of this technology. I understand that there are lots of them, but I also do feel pretty strongly that like solving these really hard problems will lead to more problems, but they are better problems, more interesting problems to solve. So I think the second order effects of this acceleration are better issues

even though there will be more of them. And that's something that does get me excited. You know the name Ted Kaczynski, right, Josh? I'm not familiar. No, you gotta fill me in. The Unabomber? Oh, yeah. Familiar with that name. Okay, do you know why the Unabomber was the Unabomber? Unfamiliar. No, please.

Please fill me in. Let's get a history lesson. Okay, so I think he had a mental disorder, something like schizophrenia or mania. I could be misremembering here. But he was a Unabomber because he had this prediction about the future, which was that the abundance of technology and the cost of technology would become so high and the cost would be so low that long-tail risks would be absolutely everywhere. And the...

Optionality for an individual to cause outsized destruction upon the planet would become increasingly available to the point where merely statistics says that something bad will happen because we are creating so much potential out of the long tail of humans.

that one human can, using technology, using infinite intelligence, ask ChatGPT or the uncensored, unhinged version of ChatGPT, how do I make an atom bomb with normal household equipment? And as we defined earlier in this podcast, the only constraint to whatever we want to do is access to intelligence, to the intelligence needed, the information needed to get what we want. And so I

I'm with you every step of the way. And I really like where this is going. And I think there's a lot of cool futures that we have. There's going to be cool technologies. We're going to be able to zip around the whole entire planet for pennies in like 10 years. It's going to be great. And I also think that most humans are good. Most humans are good people. Like 98, 99% of people are inherently good people.

But the problem is, and this is what Ted Kaczynski saw, it was that when technology is so powerful and information is so cheap and energy is so abundant, it actually only takes one person to cause outsized destruction. Have you thought about this subject? What are your thoughts or reflections? Yes. And my solution was actually derived from Palmer Luckey, who is just like the super interesting guy that I love. He's in defense tech. Uh-huh.

And he talks a lot about the war field, the dynamics of what works. He builds missiles and he builds weapons. And his thesis, and the thesis kind of since the beginning of time, is that defense is always slightly easier than offense. And this is true in software. This is true in hardware. It's always easier to defend something than it is to attack it. And in a lot of cases,

We have built these malicious tools. I mean, lots of people have guns in their house. We have nuclear bombs. We have genetic mutations. One did get away from us with COVID where, sure, maybe a handful of people did do something bad. Maybe it leaked out. Like, that's really bad. The hope is that as we...

go faster as we understand more, it will continue to be easier to defend against these bad actors than it will be for the actors to attack. And that's generally the hope. And again, it's hard to make a convincing argument to stop the progress because of the edge case of one person causing damage.

It's more interesting to continue the path forward while taking into account and being very careful about the edge cases that can harm people. I think we're seeing this a lot in AI from the leading labs like OpenAI, where they're super concerned about alignment and safety because they do understand the power of these large language models and this intelligence. And if it gets into the wrong person's hand, what can happen?

Not necessarily if it gets in the wrong person's hand, but if it's able to manipulate a large group of people into doing things, changing their minds. There's a lot of weird edge cases where we're very malleable and we are very subject to these sways of opinion. So there's always going to be that bad thing. And there's going to be lots of them. And they're going to get progressively worse, most likely. But I guess the hope is that

the trade-offs for making things better, for moving faster, will offset the downside of those few bad actors that want to use it maliciously. And there's always this double-edged sword to all progression, all technology. But you just have to hope that people are smart and are caring and are thoughtful and can work together to kind of

build something that is non-destructive. I think this is highly aligned or maybe even synonymous with Vitalik Buterin's D-AC or defensive acceleration. It also goes like decentralized acceleration. I think it's really kind of found its identity as defensive acceleration where he says, yes, I'm a fan of acceleration as concept, but I'm even more of a fan of defensive acceleration. And there are technologies that he labels that are out there that are inherently defensive rather than offensive where like,

weapons, armaments, bombs are inherently offensive, but cryptography is an inherently defensive technology. Bearer assets, Bitcoin, Ether, DeFi, inherently a defensive technology. I'm actually unfamiliar with Palmer Luckey's idea that defensive technologies are easier to create than offensive technologies. I'd love to hear like a deep dive on that. But if that is true, if those arguments are sound, then I too find myself highly skeptical

unconcerned with the Unabomber's future of the universe. Yeah, it's important to have people like Vitalik, to have people on both sides that are some people that are very like, oh, we are just going to go super fast, heads down as fast as we possibly can. And then other people that are like, wait a second, we're unlocking these, like this whole slew of attack vectors that need to be accounted for. And like, let's build solutions for that. So there needs to be stuff that happens on both sides. It does feel as if it will be easier to

defend than attack. But at the end of the day, we still have nukes. It only takes one person to start a nuclear war and that's it. So we exist in a very fragile state.

As a society, which is also why it's like, oh, like, why don't we just go to Mars and just like duplicate our intelligence there? That way we have some redundancy. Plan B. Yeah, plan B. So again, like there are ways to defend against these nukes. Well, let's just get off of the planet and let's take intelligence to this new place. So if they nuke this one, surely their nukes can't reach this next one. And you can kind of like follow this down the line of threat vectors and stuff.

figure out solutions that are kind of creative and unique to all of them. Josh, I think this conversation that we've had spawns many, many new conversations. Something that we didn't talk about this episode that I think is still highly relevant is robotics. Oh, yeah. I think there has been just like a general movement towards robotics in just very recent weeks and months. My understanding of robotics is like we're going to totally have robots everywhere

And like they're going to be humanoid robots because we have created a humanoid universe as in the form factor for navigating the universe is human. And so if we want the most effective robots, they need to be humanoid. And we can talk about like the downstream implications of that. I think there's like a nuclear episode to be done here. There is a synthetic biology episode to be done here. What would you be most interested in exploring next out of all the different doors that have been opened up?

Once we slam together the particles of energy and intelligence, and we wanna see the downstream impacts of that, what do you think are the first doors that we ought to go down here? - Those are the top three. It's energy, it is manufacturing, it is synthetic biology. Those are all amazingly impressive. The energy one is probably the least exciting, even though it's the most important.

okay, we figured out nuclear fusion fission. We can build these reactors. They supply us with a lot of energy. Cool. The manufacturing is unbelievable. The biology is even crazier. I think if you want to blow people's minds, the biology is cool because everything is downstream of biology, like the materials that we use, our own chemistry. You can recreate the biological world synthetically. That's super cool because it creates a lot of outcomes. And then the manufacturing is super cool because it's

our world will actually start to look different. That world we described where if you take your grandma and you drop her into a living room 50 years from now, the hope is that it'll look much different because we'll have robotic helpers doing things and we'll have all this autonomous transport and things that we probably can't even imagine now. So I think robotics is super cool in the sense that for the first time, we'll have computers that can coexist with us in the real world that are autonomous. Our interaction with

has only ever been static in the sense that we have a computer on our desk, we have a phone in our pocket. But they've never been able to, it's never really been two ways where there is this other person-like thing that you can actually converse with or interact with. And not only does it unlock a lot of

just quality of life improvements for us where it can do all the things that you don't want to do, but it also allows them to do all the less favorable things in life as a society that we don't want to do. Like now all of the harmful jobs, the jobs that can hurt people, they're all taken care of by robots. And now all the manufacturing that kind of sucks, that isn't super precise, that's all done by robots. And the convergence of humanoid robot and AI

alien-looking robot that does a specific narrow set of tasks, that is super cool because it opens up this whole new world. It's the autonomous transport. It's the humanoid robots. It's the swarms of drones that can deliver anything, anytime. There's a lot of really cool improvements in both of those. So I would say

manufacturing, industrial manufacturing and biology, synthetic biology in particular, are like two rabbit holes that go so deep. There is no end to them. They are black holes and they're all equally exciting in their potential. I can definitely feel your excitement through the microphone. I think maybe just one last like visual metaphor that is coming to my brain that we can leave the listeners with is something that you said earlier in this podcast is maybe what you said about Peter Thiel, which is the only thing that has really innovated in

meaningfully over the last 50 years is technology. So computers are very powerful and they've been innovating very fast. Cell phones, anything with a chip, anything electricity-wise, all technology has innovated very, very fast. Now with the arrival of AI and commoditized intelligence...

using our favorite David Deutsch quote, which is like the only constraint that we have is knowledge. And now that we have knowledge at our fingertips, we have the means of all of that innovation that technology has seen over the last 50 years. Technology can now reach back every single other industry that we haven't seen that industry and pull it forward.

simultaneously, all at once, all together, all in our lifespans and not just in our lifespans for like the next decade or so. I think we're really going to see a lot of change and that is something to definitely look forward to, maybe be scared about, maybe feel optimistic about, feel quite a lot about at the very least, no matter what. And it's going to be very exciting. And so I'm very, very interested in doing subjects around all of these episodes in the future. Josh, thanks so much for coming back on your very own podcast. Really glad to have you here. Thank you. Yeah, it's been my pleasure. It's been so much fun talking about this stuff. I really...

love chatting about it, but normally don't have many people to talk to. So thank you for listening to me. For everyone else who's listening, I appreciate it. I hope you also enjoy talking about this crazy, wacky, weird, wild future that is coming fairly quickly.

Thank you.

There's also a $9 a month ad-free podcast feed. You do not get access to the Bankless Discord. The only thing that you get is the ad-free podcast feed, but if you're tired of the ads, you can get that. Or if you become a full citizen, you can come and hang out with me, Josh, and the rest of the Bankless team in the Discord. Guys, thanks so much for bearing with us for this episode. Crypto is risky. You can lose what you put in, but nonetheless, we are headed west. This is the frontier. It's not for everyone, but we are glad you are with us on the Bankless journey. Thanks a lot. ♪